Instagram Reach Analysis Using Python

Introduction

• Instagram is one of the most popular social media applications today. People using Instagram professionally are using it for promoting their business, building a portfolio, blogging, and creating various kinds of content. As Instagram is a popular application used by millions of people with different niches, Instagram keeps changing to make itself better for the content creators and the users. But as this keeps changing, it affects the reach of our posts that affects us in the long run. So if a content creator wants to do well on Instagram in the long run, they have to look at the data of their Instagram reach. That is where the use of Data Science in social media comes in.

Goal

• In this project, Instagram Reach Analysis using Python, which will help content creators to understand how to adapt to the changes in Instagram in the long run.

Importing Libraries

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator
from sklearn.model_selection import train_test_split
from sklearn.linear_model import PassiveAggressiveRegressor

from warnings import filterwarnings
filterwarnings("ignore")

Read the dataset

df = pd.read_csv("My Instagram data.csv")

Head will give us the first five rows and columns in the dataframe

df.head()

	ID	Month	Year	category	post category	Impressions	From Home	From Hashtags	From Explore	From Other	Saves	Comments	Shares	Likes	Profile Visits	Follows	Hashtags
0	1	Jan	2021	forts	photo	5716	5050	108	72	423	144	1	250	688	21	4	#gadkille2021_maharahtra #shivray_samrajy #mar...
1	2	Jan	2021	history	video	5325	4916	20	50	323	53	9	47	786	19	0	#gadkille2021_maharahtra #shivray_samrajy #mar...
2	3	Jan	2021	facts	reel	7091	6060	50	40	939	73	2	98	898	15	0	#gadkille2021_maharahtra #shivray_samrajy #mar...
3	4	Jan	2021	shivaji maharaj	photo	4520	3491	698	194	85	25	1	93	575	5	2	#gadkille2021_maharahtra #shivray_samrajy #mar...
4	5	Jan	2021	sambhaji maharaj	photo	6257	3269	2706	350	196	22	1	353	327	23	2	#dasra #vijayadashami #dashera #gadkille2021_m...

Tail will give us the last five rows and columns in the dataframe

df.tail()

	ID	Month	Year	category	post category	Impressions	From Home	From Hashtags	From Explore	From Other	Saves	Comments	Shares	Likes	Profile Visits	Follows	Hashtags
245	246	Nov	2022	forts	video	4681	890	2379	966	10	9	6	9	382	7	6	#gadkille2021_maharahtra�#marathaempire�#shiva...
246	247	Dec	2022	history	reel	19066	428	501	17291	457	166	8	279	1097	100	22	#gadkille2021_maharahtra�#marathaempire�#shiva...
247	248	Dec	2022	sambhaji maharaj	photo	3988	340	501	2919	3	7	5	0	301	4	4	#gadkille2021_maharahtra�#marathaempire�#shiva...
248	249	Dec	2022	shivaji maharaj	photo	6124	572	2398	2710	440	73	1	61	352	5	4	#gadkille2021_maharahtra�#marathaempire�#shiva...
249	250	Dec	2022	shivaji maharaj	video	15187	1881	1989	10504	811	85	5	586	738	11	8	#gadkille2021_maharahtra�#marathaempire�#shiva...

It shows the total no of rows and columns

df.shape

(250, 17)

It shows total columns

df.columns

Index(['ID ', 'Month', 'Year', 'category', 'post category', 'Impressions',
       'From Home', 'From Hashtags', 'From Explore', 'From Other', 'Saves',
       'Comments', 'Shares', 'Likes', 'Profile Visits', 'Follows', 'Hashtags'],
      dtype='object')

df.min()

ID                                                                1
Month                                                           Apr
Year                                                           2021
category                                                      facts
post category                                                 photo
Impressions                                                    1121
From Home                                                         0
From Hashtags                                                     0
From Explore                                                      0
From Other                                                        0
Saves                                                             4
Comments                                                          0
Shares                                                            0
Likes                                                            72
Profile Visits                                                    3
Follows                                                           0
Hashtags          #dasra #vijayadashami #dashera #gadkille2021_m...
dtype: object

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 250 entries, 0 to 249
Data columns (total 17 columns):
 #   Column          Non-Null Count  Dtype 
---  ------          --------------  ----- 
 0   ID              250 non-null    int64 
 1   Month           250 non-null    object
 2   Year            250 non-null    int64 
 3   category        250 non-null    object
 4   post category   250 non-null    object
 5   Impressions     250 non-null    int64 
 6   From Home       250 non-null    int64 
 7   From Hashtags   250 non-null    int64 
 8   From Explore    250 non-null    int64 
 9   From Other      250 non-null    int64 
 10  Saves           250 non-null    int64 
 11  Comments        250 non-null    int64 
 12  Shares          250 non-null    int64 
 13  Likes           250 non-null    int64 
 14  Profile Visits  250 non-null    int64 
 15  Follows         250 non-null    int64 
 16  Hashtags        250 non-null    object
dtypes: int64(13), object(4)
memory usage: 33.3+ KB

df.describe()

	ID	Year	Impressions	From Home	From Hashtags	From Explore	From Other	Saves	Comments	Shares	Likes	Profile Visits	Follows
count	250.000000	250.000000	2.500000e+02	250.000000	250.000000	2.500000e+02	250.000000	250.000000	250.0000	250.000000	250.000000	250.000000	250.000000
mean	125.500000	2021.556000	4.662407e+04	3462.752000	1927.016000	3.921094e+04	1281.916000	310.364000	8.3120	217.508000	1993.272000	156.984000	75.560000
std	72.312977	0.497851	2.658521e+05	2765.502797	1867.021699	2.630805e+05	5089.363687	1561.463269	21.0503	1190.780387	7717.160361	1025.325745	440.689682
min	1.000000	2021.000000	1.121000e+03	0.000000	0.000000	0.000000e+00	0.000000	4.000000	0.0000	0.000000	72.000000	3.000000	0.000000
25%	63.250000	2021.000000	4.542250e+03	2032.500000	569.250000	2.985000e+02	75.250000	52.000000	2.0000	7.000000	178.000000	16.000000	4.000000
50%	125.500000	2022.000000	7.468000e+03	2749.000000	1352.000000	1.723500e+03	245.000000	95.500000	5.0000	26.500000	546.000000	29.500000	12.000000
75%	187.750000	2022.000000	1.984100e+04	4187.750000	2778.000000	1.137450e+04	736.500000	198.500000	8.0000	135.250000	1324.000000	69.750000	28.000000
max	250.000000	2022.000000	4.022772e+06	33595.000000	11817.000000	3.979419e+06	53442.000000	24225.000000	231.0000	18424.000000	112324.000000	15793.000000	6599.000000

Checking null values in datasets

df.isnull().sum()

ID                0
Month             0
Year              0
category          0
post category     0
Impressions       0
From Home         0
From Hashtags     0
From Explore      0
From Other        0
Saves             0
Comments          0
Shares            0
Likes             0
Profile Visits    0
Follows           0
Hashtags          0
dtype: int64

It gives distinct values in perticular column

df["category"].unique()

array(['forts', 'history', 'facts', 'shivaji maharaj', 'sambhaji maharaj'],
      dtype=object)

Data Visualization

Analyzing Instagram Reach

Histogram

df.hist()

array([[<AxesSubplot:title={'center':'ID '}>,
        <AxesSubplot:title={'center':'Year'}>,
        <AxesSubplot:title={'center':'Impressions'}>,
        <AxesSubplot:title={'center':'From Home'}>],
       [<AxesSubplot:title={'center':'From Hashtags'}>,
        <AxesSubplot:title={'center':'From Explore'}>,
        <AxesSubplot:title={'center':'From Other'}>,
        <AxesSubplot:title={'center':'Saves'}>],
       [<AxesSubplot:title={'center':'Comments'}>,
        <AxesSubplot:title={'center':'Shares'}>,
        <AxesSubplot:title={'center':'Likes'}>,
        <AxesSubplot:title={'center':'Profile Visits'}>],
       [<AxesSubplot:title={'center':'Follows'}>, <AxesSubplot:>,
        <AxesSubplot:>, <AxesSubplot:>]], dtype=object)

* Now let’s start with analyzing the reach of Instagram posts.look at the distribution of impressions that received from home:

plt.figure(figsize=(10, 8))
sns.set(style="darkgrid")
plt.title("Distribution of Impressions From Home")
sns.distplot(df['From Home'])
plt.show()

• The impressions get from the home section on Instagram shows how much posts reach followers. Looking at the impressions from home, I can say it’s hard to reach all followers daily. Now let’s have a look at the distribution of the impressions which are received from hashtags:

plt.figure(figsize=(10, 8))
plt.title("Distribution of Impressions From Hashtags")
sns.distplot(df['From Hashtags'])
plt.show()

• Hashtags are tools we use to categorizeposts on Instagram so that we can reach more people based on the kind of content we are creating. Looking at hashtag impressions shows that not all posts can be reached using hashtags, but many new users can be reached from hashtags. Now let’s have a look at the distribution of impressions received from the explore section of Instagram:

plt.figure(figsize=(10, 8))
plt.title("Distribution of Impressions From Explore")
sns.distplot(df['From Explore'])
plt.show()

• The explore section of Instagram is the recommendation system of Instagram. It recommends posts to the users based on their preferences and interests. By looking at the impressions we have received from the explore section, we can say that Instagram does not recommend our posts much to the users. Some posts have received a good reach from the explore section, but it’s still very low compared to the reach receive from hashtags.

Now let’s have a look at the percentage of impressions get from various sources on Instagram:

home = df["From Home"].sum()
hashtags = df["From Hashtags"].sum()
explore = df["From Explore"].sum()
other = df["From Other"].sum()

labels = ['From Home','From Hashtags','From Explore','Other']
values = [home, hashtags, explore, other]

fig = px.pie(df, values=values, names=labels, 
             title='Impressions on Instagram Posts From Various Sources', hole=0.5)
fig.show()

• So the above donut plot shows that almost 90 per cent of the reach is from home, 7.55 percent is from home, 4.2 percent is from the hashtags section, and 2.79 percent is from other sources

'''df = pd.DataFrame({
    "post category": ["forts", "history", "facts", "shivaji maharaj", "sambhaji maharaj"],
    "impressions": [100, 200, 150, 300, 600]
})'''

labels = df["category"].tolist()
values = df["Impressions"].tolist()

fig = px.pie(values=values, names=labels, 
             title='Impressions on Instagram Posts From Post Categories', hole=0.5)
fig.show()

• So the above donut plot shows that almost 51.3 per cent of the reach is from history category, 26.1 percent is from forts , 9.74 percent is from the shivaji maharaj category, and 9.08 percent is from facts category and 3.78 percent from sambhaji maharaj.

labels = df["category"].tolist()
Likes = df["Likes"].tolist()

fig = px.pie(values=Likes, names=labels, 
             title='Likes on Instagram Posts From Post Categories', hole=0.5)
fig.show()

'''df = pd.DataFrame({
    "category": ["A", "B", "C", "D", "E"],
    "count": [100, 200, 150, 300, 250]
})'''

fig = px.bar(df, x="category", y="Impressions", 
             title="Relationship between Categories and impressions",
            color="category")
fig.show()

• This bar plot is shows the history category got more impressions than others

fig = px.bar(df, x="post category", y="Impressions", 
             title="Relationship between Categories and Counts",
             color="category")
fig.show()

• This bar plot is shows the history category and post category got more impressions than others

fig = px.bar(df, x="category", y="Likes", 
             title="Relationship between Categories and likes",
             color="category")
fig.show()

fig = px.bar(df, x="category", y="Follows", 
             title="Relationship between Categories and Followers",
             color="category")
fig.show()

• from history category i got more followers that other category

pip install --upgrade scipy

Requirement already satisfied: scipy in c:\anaconda\lib\site-packages (1.10.1)
Requirement already satisfied: numpy<1.27.0,>=1.19.5 in c:\anaconda\lib\site-packages (from scipy) (1.23.5)
Note: you may need to restart the kernel to use updated packages.

import scipy
print(scipy.__version__)

1.10.1

from scipy.signal.signaltools import _centered

##### import  scipy.signal.signaltools

def _centered(arr, newsize):
    # Return the center newsize portion of the array.
    newsize = np.asarray(newsize)
    currsize = np.array(arr.shape)
    startind = (currsize - newsize) // 2
    endind = startind + newsize
    myslice = [slice(startind[k], endind[k]) for k in range(len(endind))]
    return arr[tuple(myslice)]

scipy.signal.signaltools._centered = _centered

figure = px.scatter(data_frame = df, x="Impressions",
                    y="Likes", size="Likes", trendline="ols", 
                    title = "Relationship Between Likes and Impressions")
figure.show()

• There is a linear relationship between the number of likes and the reach got on Instagram. Now let’s see the relationship between the number of comments and the number of impressions on Instagram posts:

figure = px.scatter(data_frame = df, x="Impressions",
                    y="Comments", size="Comments", trendline="ols", 
                    title = "Relationship Between Comments and Total Impressions")
figure.show()

• We can see here the 1 post have most comments, thats why this post got more impression

figure = px.scatter(data_frame = df, x="Impressions",
                    y="Shares", size="Shares", trendline="ols", 
                    title = "Relationship Between Shares and Total Impressions")
figure.show()

figure = px.scatter(data_frame = df, x="Impressions",
                    y="Saves", size="Saves", trendline="ols", 
                    title = "Relationship Between Post Saves and Total Impressions")
figure.show()

• There are also linear relationship betn Saves & Impressions.

figure = px.scatter(data_frame = df, x="Follows",
                    y="Likes", size="Likes", trendline="ols", 
                    title = "Relationship Between Post Likes and Followers Gained")
figure.show()

figure = px.scatter(data_frame = df, x="Follows",
                    y="Shares", size="Shares", trendline="ols", 
                    title = "Relationship Between Post Shares and Followers Gained")
figure.show()

figure = px.scatter(data_frame = df, x="Follows",
                    y="Comments", size="Comments", trendline="ols", 
                    title = "Relationship Between Post Likes and Followers Gained")
figure.show()

correlation = df.corr()
print(correlation["Impressions"].sort_values(ascending=False))

Impressions       1.000000
From Explore      0.999263
Profile Visits    0.988828
Follows           0.987860
Saves             0.985767
Likes             0.979050
Shares            0.975922
Comments          0.712542
From Other        0.520403
ID                0.200225
Year              0.138793
From Hashtags     0.027207
From Home         0.013366
Name: Impressions, dtype: float64

• So we can say that more saves and likes will help you get more reach on Instagram. The higher number of shares will also help you get more reach, but a low number of comments will not affect your reach either.

• Also more saves, likes and shares will help you get more followers on instagram

Analyzing Conversion Rate

• In Instagram, conversation rate means how many followers you are getting from the number of profile visits from a post.

• The formula that you can use to calculate conversion rate is (Follows/Profile Visits) * 100.

conversion_rate = (df["Follows"].sum() / df["Profile Visits"].sum()) * 100
print(conversion_rate)

48.13229373694134

conversion_rate = (df["Follows"].sum() / df["Impressions"].sum()) * 100
print(conversion_rate)

0.16206220683598807

• conversion rate of followers from impressions

conversion_rate = (df["Follows"].sum() / df["Likes"].sum()) * 100
print(conversion_rate)

3.790752090030864

• Conversion rate of follows from Likes

conversion_rate = (df["Follows"].sum() / df["Shares"].sum()) * 100
print(conversion_rate)

34.73895213049635

• Conversion rate of follows from Shares

• here we can say conversion rate into followers getting from more number of Profile Visits from the posts.

Let’s have a look at the relationship between the total profile visits and the number of followers gained from all profile visits:

figure = px.scatter(data_frame = df, x="Profile Visits",
                    y="Follows", size="Follows", trendline="ols", 
                    title = "Relationship Between Profile Visits and Followers Gained")
figure.show()

Instagram Reach Prediction Model

• Train a machine learning model to predict the reach of an Instagram post.

• Let’s split the data into training and test sets before training the model:

x = np.array(df[['Likes', 'Saves', 'Comments', 'Shares', 
                   'Profile Visits', 'Follows']])
y = np.array(df["Impressions"])
xtrain, xtest, ytrain, ytest = train_test_split(x, y, test_size=0.2, random_state=42)

We can train a machine learning model to predict the reach of an Instagram post using Python:

model = PassiveAggressiveRegressor()
model.fit(xtrain, ytrain)
model.score(xtest, ytest)

0.7463942010203388

Now let’s predict the reach of an Instagram post by giving inputs to the machine learning model:

# Features = [['Likes','Saves', 'Comments', 'Shares', 'Profile Visits', 'Follows']]
features = np.array([[282.0, 233.0, 4.0, 9.0, 165.0, 54.0]])
model.predict(features)

array([11833.40089692])

Conclusion

• So this is how you can analyze and predict the reach of Instagram posts with machine learning using Python. If a content creator wants to do well on Instagram in a long run, they have to look at the data of their Instagram reach. That is where the use of Data Science in social media comes in.